By Anna Lynn Spitzer
Irvine, Calif., Jan. 28, 2009 -- Researchers at UC Irvine’s Institute of Transportation Studies are basking in the glow of recent media attention that validates something they’ve known all along. Several online publications, including PhysOrg, Slashdot, Network World, Science Daily and Science Centric, have pointed the spotlight on “Autonet,” the vehicle-to-vehicle, autonomous traffic-information system the team is developing. While they are enjoying the publicity, the Calit2-affiliated researchers also are finding it slightly humorous that the buzz is based on a concept they began investigating nearly a decade ago.
Timing is everything, the researchers say. As wireless technology has become more common, the Autonet concept has suddenly become more relevant and the public finds the idea more realistic.
“Design, Implementation and Test of a Wireless Peer-to-Peer Network for Roadway Incident Exchange,” describes a system of ad hoc networks that allows information about traffic conditions to be passed automatically from vehicle to vehicle within a certain radius.
The paper, written by Trevor Harmon, James Marca, Pete Martini and Ray Klefstad was submitted in March 2007 to the International Journal of Vehicle Information and Communication Systems, but was not published until this month, setting off the stampede of media interest.
“It was very good fortune that [the paper] was slow to be published,” said Marca. “Enough people have seen an “OnStar” system or a live traffic update system to know this can work. That’s good but those systems are very limiting.”
The Autonet system, which would be installed in cars and roadside monitoring posts, would alleviate traffic congestion and increase driver safety by using ad hoc networks supported by common wireless technology to keep cars in contact with each other without human intervention. A car passing an accident or stuck in a traffic jam could alert other cars, allowing drivers to avoid the incident.
When it was first introduced, the concept was not well understood and was considered quite futuristic, added researcher Wenlong Jin, who in 2006 served as guest editor of a special issue of the IJVICS focused on inter-vehicle communication and intelligent transportation systems. “When we talked about this before, many people liked the idea but were kind of skeptical,” he said. “After several years, and with the continued development of wireless communication technologies, people realize it is more probable.”
Harmon attributes volume of publicity to the innovative way that two familiar technologies have been merged into a new product.
“It takes the principles of peer-to-peer networks – which are now famous thanks to Napster, BitTorrent, and the RIAA – and applies them to the problem of automobile traffic, which just about everyone cares about in one way or another,” he said. “The marrying of these two concepts in a novel way caused the paper to jump from academic literature to the mainstream press, and after one outlet picked it up, it just kind of snowballed from there.”
In its early days, the system was tested in a virtual simulator. Then, said Marca, the team collaborated with electrical engineering, computer science and database colleagues to build the prototype with off-the-shelf components, including computers connected through Bluetooth to Garmin GPS omni-directional antennae. They road-tested it on freeways and surface streets, using 802.11 b wireless networks. “That generated a lot of excitement,” he said.
The road test also laid to rest worries about a “Doppler effect” that would negatively impact communication capability when vehicles were traveling at high speed. “We showed this is not a problem with current technology,” said Jin.
Marca envisions the final product as a “hybrid system” that will incorporate all available communication options, including wireless networks and cell phone-based communication, depending on the situation.
“Some things, like a traffic jam, would be better suited to a local area connection, which is almost instantaneous,” he said. On the other hand, since the system “knows” a driver’s usual daily routes, a cell phone application could send notification before he gets in his car that there is an incident along the way. “In my mind, the end system will be both.”
Now that the academic world and the public are acquainted with the Autonet system, researchers have one more wish. “Since the paper was submitted, we haven’t gotten any additional funding,” Marca said. “Now that we’re getting press, maybe we’ll get more funding. Hopefully, we’ll get lots of money.”